top of page

LATEST PROJECTS

Project | 01

Project | Surgical Customer Effort Reduction
Reducing the amount of effort any given customer has to expend communicating and resolving complex health benefits or retirement issues is a critical initiative.  Time is an ever present constraint so getting the diagnosis right the first time is important.  This case study demonstrates how I used visual analytics to identify the most important case topics to examine from a large pool of customer interactions.

Project | 02

Project | Employee Exit Survey
Employee retention and development is a critical element of any company's strategy.  In this project I employed traditional statistical methods to understand the strength of relationship between several organizational hygiene factors such as compensation, development, and training.  Text analytics was employed to process and mine associated unstructured feedback.  Quantitative scores were then used to stratify tokenized survey comments to better understand positive/negative  employee departures and the factors most critical to positive net promoter scores.

Project | 03

Project | Customer Satisfaction Factors
Customer service agent performance is typically measured by random sampling and scoring calls using a predetermined taxonomy whose dimensions typically include personal service, knowledge, accuracy, and efficiency.  The visualization depicted here integrates the voice-of-customer satisfaction survey results segmented by the determinant factors of effort, confidence, courtesy, and efficiency.  The results are further stratified by agent ID, location, and the number of surveys associated with each agent.  Visual correlation analysis suggest that confidence and efficiency are the factors that matter most while perceptions of transaction effort are a close third.  Surprisingly, our customers don't seem to care as much about courtesy as the loosely correlated pattern suggests.  Aside from the ability to visually represent the voice of the customer as a performance metric, we can use filtering to isolate agents that perform relatively better in demanding situations such as unsuccessful transaction resolution combined with customer estimates of high effort.  Agents that score relatively better in this situation may possess more nuanced skills and techniques worthy of further examination.
I am currently open to new opportunities or freelance analytics projects.  To see more or discuss possible work let's talk >>
bottom of page